Image Classification on mini-ImageNet-Red (20% noise)
69AccuracyNCR
Evaluation Results
| Method | Links | |
|---|---|---|
| NCRBackbone=ResNet-18, Training Protocol=NCR2022.02 | 69 | |
| NCR + MixupBackbone=ResNet-18, Training Protocol=NCR, Data Augmentation=Mixup2022.02 | 68.6 | |
| StandardBackbone=ResNet-18, Training Protocol=Oracle: Clean Subset2022.02 | 68.6 | |
| MixupBackbone=ResNet-18, Training Protocol=Oracle: Clean Subset, Data Augmentation=Mixup2022.02 | 68.6 | |
| Label smoothing + MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 68.3 | |
| Label smoothingBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 68.2 | |
| MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 67.6 | |
| BootstrapBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 67.4 | |
| StandardBackbone=ResNet-18, Training Protocol=Baseline2022.02 | 66.9 | |
| Bootstrap + MixupBackbone=ResNet-18, Training Protocol=Baseline, Data Augmentation=Mixup2022.02 | 66.7 | |
| MOITBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 63.1 | |
| PropMixself-supervised learning=true2022.09 | 61.24 | |
| InstanceGM-SSself-supervised learning=true2022.09 | 60.89 | |
| InstanceGMself-supervised learning=false2022.09 | 58.38 | |
| ELRBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 57.4 | |
| D-MixBackbone=ResNet-18, Training Protocol=Other Works2022.02 | 55.8 | |
| FaMUSself-supervised learning=false2022.09 | 51.42 | |
| MentorMixself-supervised learning=false2022.09 | 51.02 | |
| DivideMixself-supervised learning=false2022.09 | 50.96 | |
| MixUpself-supervised learning=false2022.09 | 49.1 | |
| CEself-supervised learning=false2022.09 | 47.36 |